Detection of Malicious Social Bots with reinforcement learning technique with URL Features in Twitter Network with KNN in comparison with RNN

Aim: The objective of the analysis is the Real time identification of spiteful Social Bots Utilize reinforcement learning technique with URL aspects in Twitter networking with Novel K-Nearest Neighbours in comparison with Recurrent Neural Network. Materials and Methods: Realtime noticing of destructive general bots operates reinforcement knowledge technique with uniform resource locator factor in Tweet web work with K-Nearest Neighbours in comparison with Recurrent Neural Network. In this case, pre-test power analysis was done with 80% and the test size for the two groups are 20 and each group having the size of 10. Result: Real-time classification of suspicious spammers in the Online social network using reinforcement learning technique method and Webpage capabilities with K-Nearest Neighbours with the accuracy 93.89% and then compared with Recurrent Neural Network with the accuracy of 92.11% respectively. There is a statistical significance variance in the precision for two methods 0.007(p<0.05) with the use of independent sample test. Conclusion: To conclude that KNN algorithm performs significantly better than the RNN algorithm.